Is gcc allowed to eliminate writes in a destructor?

[Avi Kivity]

On 06/10/2015 10:16 AM, Norbert Lange wrote:
> Thanks for all the answers, now lets get to the next level =)
> I seen this code from one of my colleagues, and wonder wether this is valid C++:
> ~CMyclass()
> {
>     foo(this);
> }
>
> To quote the standard:
> "The lifetime of an object of type T ends when:
> * if T is a class type with a non-trivial destructor (12.4), the
> destructor call starts, or
> * the storage which the object occupies is reused or released."
>
> So you are passing a pointer to an undead object, I can expect this
> could yield interesting results in some contexts, but since you still
> have to access the object members during the destruction there isnt
> much a compiler could mess up?
> But I`d be interested what should/could happen if derived destructors
> already did their work, and foo tries to use RTTI (dynamic_cast,
> typeid). I suspect this is undefined behaviour anyway given that these
> could throw, and similarly dangerous as accessing virtual functions
> that might have overrides in the derived classes.
>
>

It's perfectly valid and very common.  You shouldn't access this after 
destruction is complete, but calling foo() during the destructor is fine.